Tandem duplication of the epidermal growth factor receptor tyrosine kinase and calcium...

Tandem duplication of the epidermal growth factor receptor tyrosine kinaseand calcium internalization domains in A-172 glioma cells

Robert A Fenstermaker1,2, Michael J Ciesielski1 and Gregory J Castiglia2

1Department of Neurosurgery, Roswell Park Cancer Institute, Elm and Carlton Streets, Bu�alo, New York 14263; and2Department of Neurosurgery, SUNY at Bu�alo School of Medicine and Biomedical Sciences, 3 Gates Circle, Bu�alo, New York14209, USA

Ampli®cation and rearrangement of the epidermalgrowth factor receptor (EGFR) gene occur frequentlyin malignant gliomas. Rearrangement may also lead tothe expression of potentially oncogenic EGFR deletionmutants. Data presented here indicate the existence of a190 kDa mutant form of the EGFR in A-172 gliomacells that is substantially di�erent from the deletionmutants characterized previously. The EGFR-like pro-tein is expressed along with the 170 kDa wild typeEGFR. It is detectable with antibodies to bothextracellular and intracellular regions of the EGFR,but is not crossreactive with other HER-family members.The wild type and mutant receptors undergo phosphor-ylation in response to treatment with TGFa and areassociated with expression of both 10.5 kb and 11.5 kbEGFR-related transcripts. Combined reverse transcrip-tion-polymerase chain reaction (RT±PCR) identi®es aunique transcript in A-172 cells that encodes an in-frame,tandem duplication of both tyrosine kinase and calciuminternalization (TK/CAIN) domains (exons 18 through26). The duplication of these domains is associated witha speci®c genomic rearrangement between potentialv-myb and c-myb consensus binding sites within introns26 and 17 of the EGFR gene resulting in the formationof a chimeric intron.

Keywords: epidermal growth factor receptor; mutation;genetic recombination; glioma; intron; oncogene

Introduction

Malignant gliomas are the most common primarybrain tumors. The prognosis for patients with thesetumors remains extremely poor (Mahaley, 1991). As aresult of multistep tumor progression, malignantgliomas are genetically heterogeneous tumors (Louisand Gusella, 1995). The most consistent geneticalterations identi®ed to date include p53 abnormalities(Bogler et al., 1995; Rasheed et al., 1994), alterations inthe p16/CDKN2/pRb pathway (Ueki et al., 1996) andEGFR gene ampli®cation and rearrangement (Liber-mann et al., 1985; Wong et al., 1987; Humphrey et al.,1988; Bigner et al., 1990). Levels of EGFR in gliomasare proportional to tumor invasiveness (Lund-Johansenet al., 1990) and overexpression of wild type EGFR hasbeen shown to induce ligand-dependent cellulartransformation of astrocytes, indicating a direct role

for the receptor in oncogenesis (Frisa et al., 1996;O'Rourke et al., 1997).The full-length (wild type) EGFR requires binding

of ligand for activation (Carpenter and Cohen, 1979).Presumably, transcription of multiple gene copies percell produces higher EGFR expression in malignantglioma cells than in normal astrocytes, with increasedsensitivity of the cell to available ligand. Hence,ectodomains of the other HER-family proteins havebeen shown to act as dominant negative factors byforming inactive heterodimers with EGFR. This leadsto inhibition of anchorage-independent growth ofhuman U87 glioma cells (O'Rourke et al., 1997).Similarly, di�erential splicing of the EGFR transcriptin normal rat liver produced a secreted 95 ± 100 kDaform of EGFR consisting of the extracellular portionof the molecule which may also act as a negativeregulator (Basu et al., 1989; Petch et al., 1990). Hence,there is abundant evidence implicating EGFR intumorigenesis of malignant gliomas, particularlyglioblastoma multiforme.EGFR ampli®cation occurs in up to 50% of

malignant gliomas with gene rearrangement inapproximately one-half of those tumors leading tofrequent co-expression of both wild type and mutantreceptors (Libermann et al., 1985; Wong et al., 1987;Humphrey et al., 1988; Bigner et al., 1990; Ekstrand etal., 1991). Expression of these EGFR mutants has beenreported to be associated with a poor prognosis(Schlegel et al., 1994). In addition, such mutants arecapable of oncogenic transformation in the absence ofligand (Batra et al., 1995; Moscatello et al., 1996).The EGFR (c-erbB-1) gene encodes a cell surface

receptor that is structurally similar to the viral erbB-1gene product (Downward et al., 1984). Proviralinsertion within intron 14 results in separation of theligand-binding domain of this receptor from itstransmembrane and cytoplasmic regions (Raines etal., 1985). The resulting N-terminally truncatedmolecule produces tissue-speci®c tumorigenesis (i.e.erythroleukemia). In contrast, certain mutations oferbB-1 occurring in the C-terminal region producesarcomas (Yamamoto et al., 1983). Such mutationsinclude C-terminal truncation, point mutations andinternal deletions within the tyrosine kinase domain(Pelley et al., 1988). Therefore, the erbB-1 gene canbecome oncogenic by way of a number of di�erentstructural alterations.Tumorigenic activation of the EGFR in human

gliomas appears to arise primarily as a result ofdeletion mutation. In addition to the 170 kDa EGFRpresent in many gliomas, certain mutant EGFR arisefrom speci®c gene arrangements and internal deletions

Correspondence: RA FenstermakerReceived 11 March 1998; revised 26 May 1998; accepted 27 May1998

Oncogene (1998) 16, 3435 ± 3443 1998 Stockton Press All rights reserved 0950 ± 9232/98 $12.00

http://www.stockton-press.co.uk/onc

(Humphrey et al., 1991; Wong et al., 1992). Suchmutants may contribute to tumor progression viaconstitutive receptor activation. Most contain adeletion of speci®c groups of exons encoding part ofthe extracellular portion of the EGFR molecule.Although uncommon in gliomas, one extensive N-terminal truncation has been reported (EGFRvI). Thismutation creates a molecule similar to the viral erbB-1gene product which induces malignant transformationand potent constitutive receptor activation (Halley etal., 1991; Wong et al., 1992). The EGFRvII mutationconsists of an in-frame deletion of 83 amino acids indomain IV of the extracellular region of the molecule(amino acids 520 ± 603; exons 14 and 15). This mutantreceptor remains capable of binding ligand and hasenhanced tyrosine kinase activity (Humphrey et al.,1991). EGFRvIII is the most common variant seen ingliomas occurring in at least 17% of them (Humphreyet al., 1990). It contains an in-frame loss of a largeportion of the extracellular region of the molecule(amino acids 6 ± 273; exons 2 ± 7). This leads to theexpression of a 140 kDa receptor with formation ofunique epitopes that are of interest for receptor-targeted antitumor strategies (Humphrey et al., 1990.Each of the EGFR mutants appears to result from lossof speci®c exons from the genes which encode them(Ekstrand et al., 1991; Wong et al., 1992). Thus, the190 kDa EGFR-like molecule, which is the subject ofthis report, represents a new and distinctly di�erenttype of EGFR mutant.

Results

Expression of a functional 190 kDa EGFR-like proteinin A-172 human glioma cells and its relationship to otherHER family proteins

EGFR-related 190 ± 200 kDa proteins have beenobserved previously in A-172, KE and A-1235 humanglioma cell lines (Steck et al., 1988; Panneerselvam etal., 1995). We have detected the EGFR-like species inA-172 cells using antibodies to both intracellular(Figure 1) and extracellular (data not shown) regionsof the EGFR. Hence, the p190 appears to be anEGFR-like molecule with similarities to both intracel-lular and extracellular regions of the wild type p170EGFR found in A-431 cells. In addition to the p190,A-172 cells express a 170 kDa form of EGFR that isindistinguishable from that produced by A-431carcinoma cells.The EGFR-like p190 present in A-172 cells is not

detectable by antibodies to HER2/neu, HER3 orHER4 proteins (Figure 2). Wild type HER1, HER2and HER4 proteins are all expressed in A-172 cells,although HER3 is not. HER3 is detectable in A-431cells, however. Consequently, based on size di�erencesof the various immunoreactive proteins, the EGFR-likep190 is unlikely to be either a wild type or mutantform of HER2, HER3 or HER4. Instead, the p190appears to be most closely related to the EGFR. Inaddition, treatment of serum-starved A-172 cells withTGFa induces phosphorylation of both 170 kDa and190 kDa species (Figure 3; lanes 5 and 6). Neither ofthe two EGFR forms is constitutively autophosphory-lated, however.

A-172 cells produce a unique 11.5 kb EGFR-liketranscript

Experiments with glycosylation inhibitors suggest thatthe protein cores of the 190 ± 200 kDa EGFR-likeprotein in A-172 and A-1235 glioma cells areglycosylated to a similar extent as the wild typeEGFR (Panneerselvam et al., 1995). Thus, thedi�erence in molecular weight between wild typeEGFR and the p190 is not due to di�erentialglycosylation. The di�erence in molecular weightbetween deglycosylated protein cores suggests that, ifthe p190 were encoded by a mutant EGFR transcript,approximately one kilobase of additional sequencewould be required beyond that present in the full-length 10.5 kb EGFR transcript (Panneerselvam et al.,1995).At least three separate EGFR-related transcripts

have been noted in A-431 carcinoma and in normalsyncytiotrophoblast cells including 10.0 ± 10.5 kb, 5.6 ±5.8 kb and 2.6 ± 2.8 kb forms (Ullrich et al., 1984;Ekstrand et al., 1991). In addition to the 10.5 kbspecies present in A-431 and A-172 cells, we haveidenti®ed an 11.5 kb EGFR-like mRNA species in thelatter (Figure 4). A transcript of similar size has alsobeen reported in KE glioma cells which have beenshown to express an EGFR-reactive 190 kDa protein(Steck et al., 1988). We have been unable to detect thelarger transcript in A-431 carcinoma cells (Figure 4) orin any other cell lines that do not express the p190.

Figure 1 Western blot of lysate from A-431 carcinoma and A-172 glioma cell lines. Anti-EGFR antibody 1005 (Santa Cruz) wasused to detect the intracellular region of the molecule. Similarresults were obtained with anti-EGFR antibody R-1 recognizingthe extracellular region (data not shown). Proteins were resolvedon 5% SDS±PAGE, transferred to nylon membranes, anddetected as described

EGFR tandem duplication mutantRA Fenstermaker et al

3436

Thus, expression of the larger transcript correspondsclosely to that of the p190. The *1 kb di�erence in thetwo transcripts is also compatible with RT ±PCRresults indicating duplication of a 1101 bp sequencewithin an EGFR-like transcript of A-172 cells asdescribed.

Characterization of EGFR transcript expression inA-172 cells by RT ±PCR using multipleoligodeoxyribonucleotide (ODN) pairs

To characterize EGFR-related transcripts present in A-172 cells, we used several di�erent sets of EGFR-

Figure 3 Duplicate Western blots of EGFR proteins from A-172 and A-431 cells. Cells were maintained in serum-free medium for24 h and treated for 10 min (+) with or (7) without TGFa (40 ng/ml) with lysates prepared as described. Anti-EGFR antibody1005 (Santa Cruz; lanes 1 ± 4) and anti-phosphotyrosine antibody PY69 (Santa Cruz; lanes 5 ± 8) were used for immunoprecipitation.Detection was with anti-EGFR antibody and a chemiluminescent system

Figure 2 Western blots of cellular proteins from A-431 carcinoma and A-172 glioma cells probed with antibodies to the indicatedHER-family members and detected with a chemiluminescent system


3437

speci®c ODN pairs in reverse transcription-polymerasechain reaction (RT ±PCR) reactions. All ODN pairsproduced fragments of identical size with A-172 and A-431 templates (Figure 5a ± d) corresponding to thelength of the wild type sequence (Ullrich et al., 1984).ODN pairs binding at the 5' and 3' ends of the EGFRcoding region (+187 and +3819; Figure 5b) alsoproduced a second major band in A-172 cells that was*1 kb larger than the wild type sequence. Similarly,ODN spanning the region from +3016 to +3819produced one PCR fragment (804 bp) in A-431 cellsand two PCR fragments (804 bp and 1905 bp) in A-172 cells (Figure 5c). Consistent results were obtainedwith two other ODN combinations (+2101 to +2292;Figure 5d; and +3116 to +3379, data not shown). Incontrast, no second band was detectable in either cellline using ODN to the extracellular region of theEGFR molecule (+187 to +2221; Figure 5a). These®ndings are consistent with the presence of twoseparate species of EGFR mRNA in A-172 cells(Figure 4). RT ±PCR reactions using independentlyisolated RNA samples gave identical results. Only thesmaller of the two species was ever detected in A-431cells, regardless of the ODN combination used.

Nucleotide sequence of the mutant EGFR mRNA

The DNA sequence of the 804 bp PCR fragmentderived from A-431 cells (Figure 5c) precisely matchedthe published EGFR cDNA sequence extending fromnucleotide +3016 in the tyrosine kinase domain to+3819 corresponding to the third nucleotide of theterminal codon. In contrast, the larger (1905 bp)

Figure 5 PCR reactions performed using four di�erent pairs of EGFR-speci®c ODN on A-172 and A-431 cDNA template mixes.ODN pairs and resulting PCR products correspond to: (a), the extracellular region of EGFR (nucleotides +187 to +2221); (b), thefull-length coding sequence (nucleotides +187 to +3819); (c) part of the intracellular region (nucleotides +3016 to +3819), and (d)part of the intracellular region (nucleotides +2101 to +2292). (e) The numbering system for ODN is derived from Ullrich et al.with numbers corresponding to the 5'-end of each ODN relative to the human EGFR coding strand. Solid lines indicate regions ofthe molecule ampli®ed by each ODN pair as depicted in a ± d

Figure 4 Northern blot of A-431 and A-172 poly(A)+RNAprobed with a DIG-labeled riboprobe spanning nucleotides+1348 to +1804 of the human EGFR cDNA (Ullrich et al.,1984). Detection was performed with a chemiluminescent systemas described


3438

fragment from A-172 cells (Figure 5c) contains atandem repeat of exons 18 through 26 as derived fromthe exon-intron structure of the chicken EGFR gene(Callaghan et al., 1993). This produces an in-frameduplication of the tyrosine kinase (TK) and calcium-mediated receptor internalization (CAIN) functionaldomains (Figure 6). In addition to tandem duplicationof the TK/CAIN region, ®ve nucleotide substitutionsare present within the sequenced region. Four of thesubstitutions occur within the third nucleotide of eachrepresentative codon leading to conservation of aminoacid sequence in each case. In the ®fth example, anucleotide substitution at +4080 (A?C) in the TK/CAIN-2 region encodes aspartate rather than gluta-mine. Analysis of two separate plasmids containing the1905 bp PCR fragment derived from a single PCRreaction and RT±PCR fragments from a di�erentisolate of A-172 RNA gave identical results.

Detection and sequence of intron rearrangement in A-172 genomic DNA with ODN from exons 26 and 18

Using an ODN homologous to the EGFR codingstrand near the 3'-end of exon 26 and an ODNhomologous to the non-coding strand near the 5'-endof exon 18, we ampli®ed a unique fragment from A-172 genomic DNA (Figure 7a, lane 7). No suchfragment could be detected in genomic DNA fromeither A-431 cells or normal human kidney (Figure 7a,lanes 8 and 9). In order to determine the signi®cance ofthis ®nding, we isolated wild type introns 17 and 26from human renal genomic DNA using PCR (Figure7a, lanes 3 and 6). Sequencing of the A-172-speci®cPCR fragment and comparison to wild type introns 17and 26 from human renal genomic DNA revealselements of both wild type introns incorporated intoa 615 bp chimeric intron (Figure 7c). The mutantintron contains 432 bp of intron 26 (wild type intron26=737 bp) and 183 bp of intron 17 (wild type intron17=982 bp). Analysis of the sites of recombination ofthe three introns using a mutation matrix derived fromsystematic binding a�nity measurements (Deng et al.,1996) revealed potential c-myb and v-myb consensusbinding sequences in the non-coding strands at sites ofrearrangement in introns 17 and 26 respectively (Figure7b). In both introns, recombination occurs within thecore regions of the putative binding sites. Moreover, a

potential c-myb binding site is reconstituted in thechimeric 26/17 intron.

Discussion

We and others have observed the presence of 190 ±200 kDa EGFR-related proteins in A-172 humanglioma cells and in two other glioma cell lines (Stecket al., 1988; Panneerselvam et al., 1995). Certaincharacteristics of these proteins such as glycosylationstatus have been described previously (Panneerselvam etal., 1995). Until the current study, however, geneticsequence alterations that might encode the p190 havenot been reported. RT ±PCR studies described hereindicate the presence of a mutant EGFR transcript inA-172 human glioma cells which encodes two copies ofthe TK/CAIN region (exons 18 through 26) in tandemarray. It is likely that this arises from a genomicalteration in which EGFR introns 26 and 17 haveundergone recombination. Together with identi®cationof the TK/CAIN duplication ampli®ed from separatemRNA isolates using several di�erent ODN pairs, the26/17 chimeric intron strongly suggests that genomicrearrangement between introns 17 and 26 creates amutant gene which encodes the TK/CAIN transcript.Moreover, our results may help to explain why previousstudies involving phosphopeptide mapping of the p190in KE cells failed to reveal signi®cant di�erencesbetween it and the wild type receptor (Steck et al.,1988).A detailed analysis of the chicken EGFR exon-

intron junction structure reveals that exon duplicationprobably occurred within the ligand binding region ofthe EGFR gene at some time during its history(Callaghan et al., 1993). Thus, as one step in theevolution of the EGFR gene, certain susceptibleintrons may have undergone genetic recombinationgiving rise to exon duplication. The multi-domainorganization of EGFR is present in the other HER-family receptors as well. Studies in which variousfunctional domains of EGFR and HER-2 (neu) areexchanged reveal that the ligand-binding, transmem-brane, and tyrosine kinase domains are capable offunctioning as independent modules (Lehvaslaiho etal., 1989). Thus, in malignant gliomas, where geneticinstability is common, EGFR rearrangements that

Figure 6 Structure of the TK/CAIN tandem duplication mutant cDNA with partial nucleotide sequence at boundaries of the TK/CAIN-2 region. Numbers at the top indicate nucleotide positions relative to the ATG start codon beginning at +187 (Ullrich et al.,1984) and are modi®ed to account for the duplicated TK/CAIN sequence. Conservative substitutions occur at nucleotides +3168(C?T) and +3345 (T?C) in the TK/CAIN-1 region and at +3648 (A?G) and +4269 (C?T) in the TK/CAIN-2 region. Anadditional substitution at +4080 (A?C) in TK/CAIN-2 leads to a glutamine to aspartate change.


3439

involve either loss or duplication of modular groupsof exons is not entirely surprising. In addition toduplication of the TK/CAIN functional domains,other regions of the EGFR, including those that arethe site of deletion mutation, can undergo tandemduplication (our unpublished observations). If suchmutants are oncogenic, tumor cells expressing themmay have a distinct survival advantage.The site of rearrangement in intron 17 consists of a

potential c-myb binding site (87.4% homology bymutation matrix) with 100% conservation of the corebinding motif (Deng et al., 1996). In addition, a v-mybconsensus sequence (89.2% homology by matrix) ispresent at the site of rearrangement in intron 26 with87.6% homology in the core. In both introns,rearrangement results from recombination within thecore regions leading to reconstitution of a potential c-myb binding sequence in the chimeric 26/17 intron(87.2% homology by matrix) with 100% homology inthe core. Although primarily a transcriptional regu-lator, the c-myb protein has been implicated in VDJrecombination at the T cell receptor delta locus wherean intact c-myb binding site is necessary for e�cientrearrangement of the minilocus substrate (Hernandez-Munain et al., 1996). Moreover, the c-myb gene itself isampli®ed and highly expressed in two of four

malignant glioma cell lines examined in one study(Welter et al., 1990). Additional studies will benecessary, however, to determine whether or not thec-myb protein is actually involved in EGFR rearrange-ment in human glioma cells.The subclass I growth factor receptors such as

EGFR have a single tyrosine kinase catalytic domain.In contrast, certain members of the subclass IIIreceptor family, such as FLT3, c-kit, c-fms andmembers of the Janus (JAK) family of receptors havetwo separate tyrosine kinase domains separated by aKI or `kinase insert' region (Agnes et al., 1994). Smallinternal tandem duplication of nucleotide sequenceswithin the juxtamembrane and TK-1 regions of theFLT3 gene have been reported in the cells of patientswith acute myeloid leukemia (Yokota et al., 1997).While these mutations may lead to constitutiveactivation and are associated with a poor prognosisfor patients, they have not yet been shown to involvefurther duplication of either the TK-1 or TK-2domains in their entirety (Horiike et al., 1997).Similarly, an in vitro mutation of the trk receptor hasbeen described which contains an inverted duplicationof the tyrosine kinase domain in head-to-tail arrange-ment (Coulier et al., 1990). Nevertheless, tandemduplication of the entire tyrosine kinase and calcium-

Figure 7 Isolation and partial structure of wild type and mutant introns. (a) PCR with genomic DNA from A-172 cells, A-431cells, and human kidney. ODN used are from exons 17 and 18 (intron 17), exons 26 and 27 (intron 26) and exons 26 and 18(chimeric intron 26/17). (b) The homology of sequences ¯anking each breakpoint to c-myb and v-myb binding sites is shown.Underlined nucleotides have strong homology to myb consensus sequences. (c) Structure of wild type introns 26 and 17 from renalgenomic DNA (a, lanes 3 and 6) and the 26/17 chimera present in A-172 cells (a, lane 7). The genomic sequences at the intron-exonboundaries and at intron breakpoints are indicated along with the corresponding nucleotide sequence derived from the mutantcDNA


3440

dependent receptor internalization domains appears tobe distinctly unusual among tumors expressing subclassI receptors.Overexpression of TGFa, which occurs in a large

percentage of malignant gliomas, may lead toautocrine-stimulated tumor cell proliferation andcellular transformation (Maruno et al., 1991; Blasbandet al., 1990). More speci®cally, ligand-dependentcellular transformation has been demonstrated inimmortalized murine glia in which the wild typeEGFR has been expressed at high levels (Frisa et al.,1996). Hence, in the presence of other critical geneticalterations, ligand-dependent glial cell transformationmay occur via overexpression of a functional EGFR.Treatment of A-172 cells with TGFa induces

tyrosine phosphorylation of wild type p170 andtandem mutant p190. Although this suggests thatboth wild type and tandem mutant receptors bindligand and undergo receptor autophosphorylation, itdoes not exclude the possibility of transphosphoryla-tion of the p190 by ligand-activated wild typereceptors. Moreover, although the TK/CAIN mutantmay be activated directly by ligand, it does not appearto be constitutively autophosphorylated. Similarly, theEGFRvIII does not exhibit enhanced basal autopho-sphorylation even though it is oncogenic (Chu et al.,1997; Huang et al., 1997). As with EGFRvIII, thetandem mutant could cause ligand-independent trans-formation despite the absence of high-level constitutiveautophosphorylation. Either or both TK domainscould phosphorylate substrates resulting in activationof a number of di�erent signal cascades. In the tandemmutant, the spatial relationship of TK-2 to itsregulatory domain is unchanged; however, TK-1 isseparated from the regulatory domain by 367 aminoacid residues. Thus, it is possible that TK-1 could beunregulated and hence oncogenic due to its spatialseparation from the regulatory domain, or as a resultof secondary conformational alterations. In addition tothe tandem duplication, a single point mutation ispresent at nucleotide +4080 located within the TK-2region of the mutant. This results in an amino acidsubstitution (glutamine to aspartate) that couldpotentially alter the function of that tyrosine kinasedomain, although the similarity of charge makes thisunlikely.Although EGFRvIII is oncogenic, the precise means

by which the aberrant signal is transduced remains asubject of investigation. In comparison to the wild typeEGFR, EGFRvIII constitutively complexes with theadapter proteins Grb2 and Shc (Montgomery et al.,1995; Moscatello et al., 1996; Chu et al., 1997).EGFRvIII also activates MAP kinase in some studies(Moscatello et al., 1996), but not in others (Chu et al.,1997). In addition, EGFRvIII activates extracellularsignal-regulated kinase (MEK) (Moscatello et al.,1996). Therefore, among other di�erences, the EGFR-vIII activates a di�erent complement of substrates thanthe wild type receptor. Thus, co-expression of wild typeEGFR and one or more of the EGFR mutants, couldhelp to maintain a transformed state by simultaneousactivation of parallel signaling pathways.There is evidence that EGFRvIII may be constitu-

tively active as a result of ine�cient receptor down-regulation, rather than by way of enhancedautophosphorylation (Chu et al., 1997; Huang et al.,

1997). Thus, unlike wild type receptors, EGFRvIIImay be unable to attenuate mitogenic signaling byreceptor internalization (Huang et al., 1997). Withinthe C-terminal region of the EGFR there is a domaindesignated CAIN which modulates receptor internali-zation (Chen et al., 1989). Deletion of this region leadsto increased transforming ability presumably viaprolongation of cell-surface receptor half-life (Wells etal., 1990). Thus, there are a number of possiblemechanisms by which deletion mutations in theEGFR could lead to propagation of an oncogenicsignal in the absence of ligand. It will be of interest todetermine the functional e�ects that duplication of theCAIN region might have on this process, as well as thee�ects of tyrosine kinase duplication on glial transfor-mation and signal transduction

Materials and methods

Cell culture

A-172 astrocytoma and A-431 carcinoma cell lines weregrown on 100 mm tissue culture plates in Dulbecco'sModi®ed Eagle Medium (DMEM) with 1500 mg/L glu-cose, 2 mM L-glutamine, and 110 mg/L sodium pyruvate,supplemented with 10% fetal bovine serum, 50 units/mlpenicillin and 50 mg/ml streptomycin. Cells were grown at378C in 5% CO2 with media changes 2 ± 3 times per week.

RNA isolation and Northern blotting

Poly(A)+ RNA was isolated from A-172 and A-431 cellsusing a batch oligo dT method (Invitrogen). RNA sampleswere resolved on formaldehyde-agarose gels and transferredto positively charged nylon membranes (MSI) using aturboblotting apparatus as recommended by the manufac-turer (Schleicher and Schuell). Blots were hybridized with aDIG-UTP-labeled riboprobe generated from pBluescript(KS7; Stratagene) containing a 456 bp BamHI ±HindIIIrestriction fragment from the human EGFR cDNA(nucleotides +1348 to +1804; Ullrich et al., 1984).Hybridization of the EGFR riboprobe was detected usingalkaline-phosphatase-conjugated, anti-DIG antibodies andthe CDP-STAR chemiluminescent substrate (Bio-Rad).

Reverse transcription (RT) and PCR analysis of EGFR cDNA

RT reactions were performed with poly(A)+ RNA andusing oligo dT and MMLV reverse transcriptase (Perkin-Elmer). RT reactions were examined for EGFR cDNAs byPCR using multiple oligodeoxyribonucleotide (ODN) pairs.ODN were prepared on an Applied Biosystems DNAsynthesizer in the Biopolymer Facility at Roswell ParkCancer Institute. All ODN used for PCR were designed witha TM of 668C. PCR fragments were generated using amixture of 1.25 units Taq polymerase (Life Technologies)and 0.075 units Pfu polymerase (Stratagene). Reactionswere incubated in a thermal cycler (Perkin Elmer) at 948Cfor 5 min, then 948C for 30 s alternating with 728C for6 min 65 cycles; then 948C for 30 s alternating with 708Cfor 6 min6®ve cycles; then 948C for 30 s alternating with688C for 6 min625 cycles. Reaction products were resolvedon 1% agarose gels and eluted using Geno-bind reagent(Clontech). Isolated fragments were tailed with Taqpolymerase and dATP and introduced into the pCRII-Topo vector (Invitrogen). Sequencing was performed usingautomated methods and homology searches were performedusing MatInspector (Quandt et al., 1995). Each plasmidinsert was sequenced at least twice.


3441

Western blotting and immunoprecipitation

Cells were lysed in RIPA bu�er (50 mM Tris-HCl, pH=7.5,0.25% sodium deoxycholate, 1% NP-40, 1 mM sodiumorthovanadate, 1 mM sodium ¯uoride, 1 mM EGTA, 1 mg/ml each of pepstatin, leupeptin and aprotinin). Lysates werepassed through a 20 guage needle 15 times to shear DNAand centrifuged to remove debris. Protein content wasmeasured by the bicinchoninic acid method (Sigma).Cellular extracts (50 mg) were boiled for 5 min and appliedto 5% SDS-polyacrylamide gels. Protein was electroblottedonto positively charged nylon membranes (MSI) and thenprobed with anti-EGFR, neu, HER3, or HER4 antibodies(Santa Cruz). Speci®c binding was detected by secondaryantibodies conjugated to alkaline phosphatase which wasthen exposed to CDP-STAR chemiluminescent substrate(Bio-Rad). Immunoprecipitations were performed with

equal amounts of protein extract from A-172 and A-431cells using either anti-EGFR (EGFR 1005; Santa Cruz) oranti-phosphotyrosine (PY69; Santa Cruz) antibodies (1 mg/ml) followed by incubation with protein-A agarose beads.Immunoprecipitates were washed three times with RIPAbu�er, boiled in 26Laemmli bu�er, run on 5% SDS-polyacrylamide gels and detected by Western blotting withanti-EGFR antibody as described above.

AcknowledgementsThe authors wish to thank Dr Hsing-Jien Kung and Dr JimJacobberger of Case Western Reserve University for theiradvice and encouragement. This project was supported byNIH grant CA 16056-21.

References

Agnes F, Shamoon B, Dina C, Rosnet O, Birnbaum D andGalibert F. (1994). Gene, 145, 283 ± 288.

Basau A, Raghunath M, Bishayee S and Das M. (1989).Mol.Cell. Biol., 9, 671 ± 677.

Batra SK, Castelino-Prabhu S, Wikstrand CJ, Zhu X,Humphrey PA, Friedman HS and Bigner DD. (1995).Cell Growth Di�er., 6, 1251 ± 1259.

Bigner SH, Humphrey PA, Wong AJ, Vogelstein B, Mark J,Friedman HS and Bigner DD. (1990). Cancer Res., 50,8017 ± 8022.

Blasband AJ, Gilligan DM, Winchell LF, Wong ST,Luetteke NC, Rogers KT and Lee DC. (1990).Oncogene, 5, 1213 ± 1221.

Bogler O, Huang HJ and Cavenee WK. (1995). Cancer Res.,55, 2746 ± 2751.

Callaghan T, Anctczak M, Flickinger T, Raines M, Myers Mand Kung H-J. (1993). Oncogene, 8, 2939 ± 2948.

Carpenter G and Cohen S. (1979). Ann. Rev. Biochem., 48,193 ± 216.

Chen WS, Lazar CS, Lund KA, Welsh JB, Chang CP,Walton GM, Der CJ, Wiley HS, Gill GN and RosenfeldMG. (1989). Cell, 59, 33 ± 43.

Chu CT, Everiss KD, Wikstrand CJ, Batra SK, Kung H-Jand Bigner DD. (1997). Biochem. J., 324, 855 ± 861.

Coulier F, Kumar R, Ernst M, Klein R, Martin-Zanca D andBarbacid M. (1990). Mol. Cell. Biol., 10, 4202 ± 4210.

Deng Q-L, Ishii S and Sarai A. (1996). Nucleic Acids Res., 24,766 ± 774.

Downward J, Yarden Y, Mayes E, Scrace G, Totty N,Stockwell P, Ullrich A, Schlessinger J and Water®eld MD.(1984). Nature, 397, 521 ± 527.

Ekstrand AJ, James CD, Cavenee WK, Seliger B, PettersonRF and Collins VP. (1991). Cancer Res., 51, 2164 ± 2172.

Frisa PS, Walter EI, Ling L, Kung H-J and Jacobberger JW.(1996). Cell Growth Di�er., 7, 223 ± 233.

Haley JD, Hsuan JJ and Water®eld MD. (1990). Oncogene,4, 273 ± 283.

Hernandez-Munain C, Lauzurica P and Krangel MS. (1996).J. Exp. Med., 183, 289 ± 293.

Horiike S, Yokota S, Nakao M, Iwai T, Sasai Y, Kaneko H,Taniwaki M, Kahsima K, Fujii H, Abe T and Misawa S.(1997). Leukemia, 11, 1442 ± 1446.

Huang H-J Su, Nagane M, Klingbeil K, Lin H, Nishikawa R,Xiang-Dong J, Huang C-M, Gill GN, Wiley HS andCavenee WK. (1997). J. Biol. Chem., 272, 2927 ± 2935.

Humphrey PA, Wong AJ, Vogelstein B, Friedman HS,Werner MH, Bigner DD and Bigner SH. (1988). CancerRes., 48, 2231 ± 2238.

Humphrey PA, Wong AJ, Vogelstein B, Zalutsky MR, FullerGN, Archer GE, Friedman HS, Kwatra MM, Bigner SHand Bigner DD. (1990). Proc. Natl. Acad. Sci. USA, 87,4207 ± 4211.

Humphrey PA, Gangarosa LM, Wong AJ, Archer GE,Lund-Johansen M, Bjerkvig R, Laerum O-D, FriedmanHS and Bigner DD. (1991). Biochem. Biophys. Res.Comm., 178, 1413 ± 1420.

Lehvaslaiho J, Lehtola L, Sistonen L and Alitalo KA. (1989).EMBO J., 8, 159 ± 166.

Liberman TA, Nusbaum HR, Razon N, Kris R, Lax I, SoreqH, Whittle N, Water®eld MD, Ullrich A and SchlessingerJ. (1985). Nature, 313, 144 ± 147.

Louis DN and Gusella JF. (1995). Trends Genet., 11, 412 ±415.

Lund-Johansen M, Bjerkvig R, Humphrey PA, Bigner SH,Bigner D and Laerum O-D. (1990). Cancer Res., 50,6039 ± 6044.

Mahaley MS. (1991). J. Neuro-Oncol., 11, 85 ± 147.Maruno M, Kovach JS, Kelly PJ and Yanagihara T. (1991).

J. Neurosurg., 75, 97 ± 102.Montgomery RB, Moscatello DK, Wong AJ, Cooper JA andStahl WL. (1995). J. Biol. Chem., 270, 30562 ± 30566.

Moscatello DK, Montgomery RB, Sundareshan P, McDanelH, Wong MY andWong AJ. (1996). Oncogene, 13, 85 ± 96.

O'Rourke DM, Qian X, Zhang H-T, Davis JG, Nute E,Meinkoth J and Greene MI. (1997). Proc. Natl. Acad. Sci.USA, 94, 3250 ± 3255.

Panneerselvam K, Kanakaraj P, Raj S, Das M and BishayeeS. (1995). Eur. J. Biochem., 230, 951 ± 957.

Pelley RJ, Moscovici C, Hughes S and Kung H-J. (1988). J.Virol., 54, 304 ± 310.

Petch LA, Harris J, Raymond VW, Blasband A, Lee DC andEarp HS. (1990). Mol. Cell. Biol., 10, 2973 ± 2982.

Quandt K, Frech K, Karas H, Wingender E and Werner T.(1995). Nucleic Acids Res., 23, 4878 ± 4884.

Raines MA, Lewis WG, Crittenden LB and Kung H-J.(1985). Proc. Natl. Acad. Sci. USA, 82, 2287 ± 2291.

Rasheed BK, McLendon RE, Herndon JE, Friedman HS,Friedman AH, Bigner DD and Bigner SH. (1994). CancerRes., 54, 1324 ± 1330.

Schlegel J, Merdes A, Stumm G, Albert FK, Forsting M,Hynes M and Kiessling M. (1994). Int. J. Cancer, 56, 72 ±77.

Steck PA, Lee P, Hung M-C and Yung WKA. (1988). CancerRes., 48, 5433 ± 5439.

Ueki K, Ono Y, Henson JW, E®rd JT, von Deimling A andLouis DN. (1996). Cancer Res., 56, 150 ± 153.

Ullrich A, Coussens L, Hay¯ick JS, Dull TJ, Gray A, TamAW, Lee J, Yarden Y, Libermann TA, Schlessinger J,Downward J, Mayes ELV, Whittle N, Water®eld MD andSeeburg PH. (1984). Nature (London), 309, 418 ± 425.

Wells A, Welsh JB, Lazar CS, Wiley HS, Gill GN andRosenfeld MG. (1990). Science, 247, 962 ± 964.

Welter C, Henn W, Theisinger B, Fischer H, Zang KD andBlin N. (1990). Cancer Ltrs., 52, 57 ± 62.


3442

Wong AJ, Bigner SH, Bigner DD, Kinzler KW, Hamilton SRand Vogelstein B. (1987). Proc. Natl. Acad. Sci. USA, 84,6899 ± 6903.

Wong AJ, Ruppert JM, Bigner SH, Grzeschik CH,Humphrey PA, Bigner DS and Vogelstein B. (1992).Proc. Natl. Acad. Sci. USA, 89, 2965 ± 2969.

Yamomoto T, Hihara H, Nishida T, Kawai S andToyoshima K. (1983). Cell, 34, 225 ± 232.

Yokota S, Kiyoi H, Nakao M, Iwai T, Misawa S, Okuda T,Sonoda Y, Abe T, Kahsima K, Matsuo Y and Naoe T.(1997). Leukemia, 11, 1605 ± 1609.


3443

Tandem duplication of the epidermal growth factor receptor tyrosine kinase and calcium...

Documents

Transcript of Tandem duplication of the epidermal growth factor receptor tyrosine kinase and calcium...